Friday, September 23, 2005

Double-Blind

A piece of research done by colleagues in the politics department here has been reported on the Guardian's website. According to the research, when exams are marked according to a double-blind system (where two people mark separately, without being aware of what the other marker has awarded the paper) there is a large discrepancy than when second markers are aware of the mark awarded by the first. It may all seem a bit arcane, but it undermines the idea that the more cost-efficient approaches to examining are just as accurate in ascertaining the worth of a student's work.

No comments: